Leap Motion Hello world:

sudo leapd
python3 plot_hand.py

Myo Hello World:

python3 myo_raw.py

Plots Myo data, note the Myo sensor with the LED is the 4th graph from the top.
You can confirm this by sliding your fingers under each sensor and watching each graph change.

Generating a python wrapper for the leap motion.

To see model development and an overview of the decisions I made while making models see modelProgress.md.

For model development, see the Notebooks folder. If they don't render in GitHub the first time, keep retrying until they do.

NeuroLeap.py - Contains functions used for getting, plotting, saving and interpreting data from the Myo and Leap.

make_dataset.py - Uses data_gather.py to read data from a single Myo and Leap motion until a time limit is reached, then saves the Myo and Leap data into one CSV file. The Leap motion data is plotted in the main thread to monitor the quality of the Leap Motion data.

data_gather.py - Used by other scripts to connect and gather data from a single Myo and pass this data back to the main script using a multiprocessing array.

plot_hand.py - Live plots 22 points of the hand gathered from the Leap Motion using Matplotlib.

animate_saved.py - Animates gathered LeapMotion data, useful to see exactly what gestures were performed in each dataset.

myo_examp.py - An example program that by default, plots 50Hz rectified bandpass filtered Myo data. Shows how to gather myo data in a seperate thread and plot it using pygame.

plot_angle_from_rot.py - Generates angles, by creating a rotation matrix from basis vectors. This is a work in progress.

1. Fix your leap motion in place, e.g. by using bluetack, draw an outline around it, just incase.

2. Put the Myo armband on the thickest part of your forearm, with the LED bar pointing down towards your fingers.

3. Wait 5 minutes for the Myo to warm up.

4. Open make_dataset.py, set the filter settings for the Myo, the dataset name and the time that data should be gathered. Too long and your fingers will fatigue, affecting the gathered data. When done, a CSV of LeapMyo data will be saved in the project root folder.

5. Use animate_saved.py to watch this CSV back and see if the Leap glitches. If it's fine, use it with a notebook to make a model.

6. Provide predict_hand_points.py with a keras model that outputs the relative position of the fingers to the palm.